DESCRIPTION (from abstract): There are 3.5 million Americans with low vision, most of whom have reading difficulty. Macular degeneration is the leading cause, and often results in a loss of central vision. The reading deficits of people with central-field loss is a major, unsolved health problem, and is the primary focus of this proposal. Dr. Legge and his colleagues have developed an ideal-observer model of reading that provides a theoretical basis for understanding how central-field loss affects reading. The concept of the visual span plays a key role in the model. The visual span is the number of letters that are identified in a single glance during reading. The normal visual span is limited to about 10 letters in width. This small visual span is further diminished or distorted as a consequence of eye disease. The main hypothesis underlying this proposal is that the properties of the visual span can account for much (and perhaps all) of the unexplained reading deficits from central-field loss. In four series of experiments, Dr. Legge proposes to test specific hypotheses that account for reading deficits in terms of (i) spatiotemporal properties of letter recognition across the visual span; (ii) letter interactions within the visual span; (iii) global properties of the visual span (shape and size); and (iv) the influence of linguistic and motor constraints. The research plan also includes development of two new clinical tests, one for assessing contrast deficits in reading, and the other for measuring Braille reading speed.